When I published my article, “The Middle Way for UK AI – Ethical Implementation for the UK’s AI Opportunities Action Plan” last year, I argued that the UK is uniquely positioned to chart a middle course between the unregulated frontier of the United States and the prescriptive rigidity of the European Union. Through a principles-based, pro-innovation approach that does not sacrifice human dignity at the altar of speed. As well as, by leveraging the UK’s AI Opportunities Action Plan, built on laying foundations, driving adoption, and securing the future, we can harness AI for economic growth while embedding vital ethical safeguards.
This argument extends beyond the UK. The governance of artificial intelligence should be a synthesis, a middle way, that balances the transformative economic potential of AI with a moral imperative to ensure its benefits are broadly shared. Nowhere is this approach more urgently needed, and uniquely achievable, than across the Commonwealth.
The Uncomfortable Reality of the AI Divide
The numbers are stark. According to the World Bank’s 2025 report on “Who on Earth Is Using Generative AI? Global Trends and Shifts in 2025”, 24% of internet users in high-income countries use generative AI, compared with just 0.7% in low-income countries and 5.8% across developing countries. By contrast, Microsoft’s AI Diffusion Report and Stanford HAI’s AI Index put Singapore’s GenAI user penetration at 75.2%, while Canadian adoption research from Statistics Canada and KPMG Canada places Canada, a high-income Commonwealth member, at 31.6%. This is not merely a gap. It is a chasm, one that threatens to institutionalize a permanent digital underclass across the Commonwealth’s 56 member states.
A recent roundtable that I led and chaired on “Bridging the AI Divide in the Commonwealth” at the Commonwealth Secretariat, Marlborough House on 15 June 2026, in consultation with the Commonwealth Secretariat, brought this reality into sharp relief. I also provided a report with key findings and recommendations from this roundtable discussion to the Commonwealth Secretariat. Participants from government, industry, and diplomatic missions across nations such as: Fiji, Seychelles, Pakistan, and Rwanda, described a multi-layered divide that extends far beyond connectivity. It encompasses generational gaps, where rapid adoption of AI tools by younger populations leaves older populations at risk of exclusion. As well as, geographic isolation, with Archipelagos like Fiji, that face severe hurdles delivering uniform digital infrastructure across numerous islands. It also encompasses institutional fragility, where political shifts undermine the sustained strategy development that AI governance requires. Given the speed at which AI technology is evolving, targeted and locally adapted strategies are urgently needed to ensure this gap does not widen further.
Governance Frameworks: Learning from Singapore and the UK
The insights from the above-mentioned roundtable, fed directly into a strategic briefing that I prepared and delivered for Commonwealth policymakers on “The Governance of Agentic AI in the Commonwealth” for the Commonwealth – Singapore Third Country Training Programme on Digital Governance and AI Adoption on 6 August 2026 in collaboration with the Commonwealth Secretariat and the Civil Service College Singapore.
Singapore’s Model AI Governance Framework for Agentic AI (launched in January 2026 and updated in May) offers a compelling blueprint. It is built on four dimensions: assessing and bounding risks, meaningful human accountability, end-user responsibility, and technical controls. The framework is sector-agnostic and principle-based, capability-based in its risk framing, and whole-of-nation in its approach. Critically, it was developed iteratively, Version 1.0 was implemented in six months and refined through live deployment rather than theoretical debate. This reflects a pragmatic core philosophy: do not wait for perfect regulation while technology is three generations ahead.
The UK’s approach offers complementary lessons. The UK’s AI Opportunities Action Plan lays foundations through a £100 million National Data Library and the Isambard-AI supercomputer. It drives adoption through a “Scan > Pilot > Scale” approach across public services. It secures the future through a Sovereign AI Fund with £500 million in backing. Additionally, the UK’s regulatory structure, combining ICO’s enforcement powers under GDPR and Ofcom’s Online Safety Act, demonstrates that innovation and accountability can coexist.
For the Commonwealth, synthesizing Singapore’s agile, iterative governance with the UK’s comprehensive assurance framework offers a viable model. I proposed a “Responsible AI Framework for the Commonwealth” within the above-mentioned strategic briefing for Commonwealth policymakers, structured around governance, strategy, policy, and software development life cycle that would provide member states with a practical toolkit for implementation.
A Two-Track Strategy for a Diverse Commonwealth
The Commonwealth is not a monolith. It spans a wide spectrum of development, from high AI-readiness nations such as the UK to Small Island Developing States with limited connectivity and infrastructure. A one-size-fits-all approach to AI governance risks leaving smaller states behind.
To address this, I proposed a “Two-Track Strategy for the Commonwealth”, in the above-mentioned strategic briefing incorporating the “Responsible AI Framework for the Commonwealth”, that leverages the Commonwealth’s diversity as a key strength:
- Track 1 (Frontier and Advanced Economies): Should focus on scaling women-led AI innovation, AI governance leadership, and cross-border data partnerships. Priority should be given to harmonized regulatory frameworks, advanced alignment testing, and interoperable legal frameworks for agentic AI.
- Track 2 (Small and Developing States): Should focus on foundational digital access, AI upskilling, low-code/ no-code entrepreneurship platforms, and affordable connectivity. Priority should be given to building the essential infrastructure required to ensure meaningful participation in the AI ecosystem.
This framework is not a tiered system of worth, but a recognition that the path to AI sovereignty depends on a nation’s starting point. Member countries do not need to take identical actions, but rather to ensure that they have the same focus and direction.
Frugal AI and Mobile-First Initiatives
If the above-mentioned “Two-Track Strategy for the Commonwealth” provides the framework, frugal AI provides the method. Competing with the US or China on frontier AI models is neither feasible nor desirable for most Commonwealth members. The goal should not be to build large language models from scratch, a pursuit requiring compute power and energy that few nations can afford. Instead, the focus should be on fine-tune existing models, adapting them to local contexts, and deploying AI that is lightweight, low-cost and energy efficient.
A mobile-first approach should be the key driving principle since significant proportions of populations in developing states access the internet primarily through smartphones. AI solutions must be designed for low bandwidth, low cost, and mobile accessibility. The mobile industry can provide compute capabilities through handsets, enabling edge computing that reduces reliance on centralised data centres.
Public-private partnerships should be prioritized to upskill individuals in AI by working with local partners to disseminate technical knowledge in accessible terms. The Government of Pakistan recently formed such a partnership to expand AI and digital skills, ultimately driving the growth of the digital economy. A strong example of effective collaboration among Commonwealth countries in AI was the roundtable on “Leveraging AI for Better and Inclusive Healthcare Delivery – UK and Pakistan,” hosted by the Pakistan High Commission in London on September 24, 2025. I had the privilege of facilitating the session and presenting a strategic briefing on “Responsible AI and Healthcare.” Attendees included UK and Pakistan Health Ministers, policymakers, academics, AI experts, doctors, healthtech entrepreneurs, representative from NADRA, CEO of Pakistan Software Export Board (PSEB) and the High Commissioner of Pakistan to the UK. Following the event, a report outlining key findings and recommendations was submitted by the Pakistan High Commission in London, UK to the Government of Pakistan. These recommendations are currently being implemented by NADRA in Pakistan.
In agriculture, farmers should leverage crop photo diagnostics to provide immediate, low-cost solutions tailored to local constraints. In healthcare, frugal AI solutions such as AI-powered virtual doctors, can effectively expand access to healthcare in remote communities across the Commonwealth. Finally, energy efficiency is not merely a technical consideration; it is a matter of climate justice. The carbon footprint implications of AI are massive, and even countries without data centres will face carbon taxes. The question of how to apportion responsibility fairly is one that the Commonwealth, with its diverse membership, is uniquely positioned to address.
The NextGen Innovation Challenge 2026, organized by the National Board for Technology Incubation (NBTI), Government of Nigeria, and UKALD, , is a prime example of encouraging frugal AI, mobile-first initiatives, and energy-efficient solutions. Having served as a judge for Nigeria’s leading innovation platform, I have witnessed first-hand the practical need for these innovations, particularly in healthcare, education, agriculture, and climate resilience for rural populations, and their potential to scale across the Commonwealth. The NextGen Innovation Challenge 2026 International Grand Finale will take place on 9 October 2026 at ExCeL London, UK. Attendees will include Commonwealth government representatives, policymakers, investors, entrepreneurs, industry leaders, trade bodies, accelerators, and media.
The Gender Dimension: From Digital Literacy to AI Leadership
The AI divide is not gender neutral. Across the Commonwealth, women and underrepresented groups face a double bind of lower access to digital tools and fewer opportunities to develop AI skills. Algorithmic bias towards women in large language models that are utilized by Micro, Small and Medium Enterprises (MSMEs), particularly in areas such as recruitment, funding assessments, credit scoring, loan applications etc., further amplify these existing inequalities for underrepresented and underserved communities.
Moving from basic digital literacy to AI leadership requires a multi-pronged approach, one that addresses infrastructure, education, and cultural barriers. It requires targeted mentorship, professional certifications, such as ISO 42001 (the international standard for AI management systems), intergenerational programmes for policy leaders, and community-led digital literacy initiatives.
I can recommend initiatives such as a “Commonwealth AI Inclusion Fund,’ that focuses on a public-private partnership to provide seed capital, technical assistance, and mentorship for women-led and underrepresented communities-led AI ventures. The fund should prioritize entrepreneurs from Small Island Developing States and underserved regions, with clear gender and inclusion targets to de-risk early-stage investment.
But funding alone is insufficient. The “human in the loop” principle must be embedded in all AI initiatives. AI literacy must emphasize critical evaluation to identify AI hallucinations and incorrect outputs, particularly for younger generations. Preserving cultural identity and human dignity in an increasingly AI-driven world is not a soft consideration but remains a central governance imperative.
Unlocking the $2 Trillion Opportunity
The economic prize is immense. Agentic AI, systems that act autonomously to execute complex tasks, could unlock a $2 trillion digital trade by 2030. The Commonwealth Connectivity Agenda, guided by United Nations Commission on International Trade Law (UNCITRAL) principles, aims to harmonize digital trade regulations, and enable this potential.
However, without unified legal recognition of autonomous agents, this economic potential will fracture into incompatible domestic silos. The liability gap, the absence of clear legal frameworks governing who pays when an autonomous agent makes a mistake, must be closed. Machine-to-machine payment infrastructure must be upgraded to allow agents to execute micro-transactions and settle contracts autonomously. Model laws on electronic transferable records must create secure, legally binding digital corridors for paperless trade.
Majority of businesses within the Commonwealth are Micro, Small and Medium Enterprises MSMEs). Digital skills training would significantly benefit these businesses, driving hob creation and expanding access to finance. Furthermore, public-private partnerships would help governments by enabling them to work with technology platforms to effectively scale AI literacy.
The Commonwealth Advantage: Compute and Context
The Commonwealth’s unique advantage lies in its combination of compute and connectivity, provided by the Commonwealth Artificial Intelligence Consortium (CAIC) and tech partners, alongside local context and competency provided by the 56 Member States. Compute and connectivity alone lead to structural divergence; true resilience can only be achieved when global tech is supported by local context and competency within the 56 Member States, collaborating effectively to solve common problems and benefiting from scale across the Commonwealth.
The Commonwealth isn’t just a group of nations, it is a valuable asset, an incredible global network of cross-sector, purpose-driven innovators, and a competitive advantage. In a fractured world, the Commonwealth stands apart – built on trust, partnerships, and a shared commitment to open trade, the rule of law and fairness. With 56 pragmatic, pro-trade nations, 2.7 billion people with 60% under the age of 30, spanning 5 global regions, from G7 and G20 to Africa, ASEAN, the Caribbean, and the Pacific. A 21% trade advantage and Commonwealth laws underpinning 80% of global trade and 60% of global finance. A combined GDP of $14.2 trillion in 2022 that is projected to reach $20 trillion by 2029. Shared language, heritage, legal systems, and standards reduce friction and risk. Opportunities to unlock digital trade, modernize infrastructure and cut barriers, leverage technology and innovation, connect capital to investible projects and access the heart of the global economy.
A Call to Action
The path forward requires sustained collaboration, political will, and investment. The Commonwealth has a unique opportunity to reach a consensus on a “common responsible, ethical and equitable AI framework” based on the above-mentioned “Two-Track Strategy for the Commonwealth” that builds on the four pillars of responsible and ethical AI, data sovereignty, equitable development, and workforce resilience.
However, frameworks alone are not enough. What is needed is a shared commitment to a balanced approach or “the middle way” that recognizes AI governance as both a technical and ethical endeavour. It must pursue innovation that serves human flourishing, regulation that enables growth, and a digital future that is inclusive by design. The core values of the 2013 Commonwealth Charter must continue to guide efforts amidst challenges like deepfakes and algorithmic bias.
The 2013 Commonwealth Charter’s ethical framework remains as relevant today as when it was signed. In an age of agentic AI, deepfakes, and algorithmic bias, the principles of human dignity, equality, and shared prosperity must guide every initiative.
While the AI divide in the Commonwealth spans generational, geographic, infrastructural, educational, and financial barriers, solutions like frugal AI, mobile-first design, and community-led digital literacy offer viable pathways for even the most resource-constrained members to participate in the AI revolution. Rwanda’s investment in AI policy and digital infrastructure illustrates that political will can transform a developing nation.
The Commonwealth is uniquely positioned to bring together voices from government, industry, and civil society, to lead this effort in a trusted environment by addressing the most pressing challenges of our time. The question is not whether we can bridge the AI divide. The question is whether we will choose to.
The middle way is not a compromise. It is a strategic imperative to ensure that the Commonwealth – diverse, collaborative, and committed to leaving no one behind – is uniquely positioned to lead.
